diff --git a/src/common/low_precision_transformations/include/low_precision/common/fake_quantize_dequantization.hpp b/src/common/low_precision_transformations/include/low_precision/common/fake_quantize_dequantization.hpp
index 48ec252ece3729..29e3805f8b5b94 100644
--- a/src/common/low_precision_transformations/include/low_precision/common/fake_quantize_dequantization.hpp
+++ b/src/common/low_precision_transformations/include/low_precision/common/fake_quantize_dequantization.hpp
@@ -37,16 +37,16 @@ class LP_TRANSFORMATIONS_API FakeQuantizeDequantization {
 
     static bool checkElementwise(const std::shared_ptr<ngraph::Node>& elementwise);
 
-    static bool checkShape(const std::shared_ptr<ngraph::Node>& elementwise) noexcept;
+    static bool checkShape(const std::shared_ptr<ngraph::Node>& elementwise);
 
     static int fillDequantizationParams(
         const std::shared_ptr<ngraph::Node>& elementwise,
         std::shared_ptr<ngraph::opset1::Convert>& convert,
-        std::shared_ptr<ngraph::opset1::Constant>& constant) noexcept;
+        std::shared_ptr<ngraph::opset1::Constant>& constant);
 
     static int fillDequantizationParams(
         const std::shared_ptr<ngraph::Node>& elementwise,
-        std::shared_ptr<ngraph::opset1::Constant>& constant) noexcept;
+        std::shared_ptr<ngraph::opset1::Constant>& constant);
 
     Output<Node> data;
     std::shared_ptr<opset1::Convert> convert;
diff --git a/src/common/low_precision_transformations/include/low_precision/concat.hpp b/src/common/low_precision_transformations/include/low_precision/concat.hpp
index 3ef2c466a81928..724389b8a266a0 100644
--- a/src/common/low_precision_transformations/include/low_precision/concat.hpp
+++ b/src/common/low_precision_transformations/include/low_precision/concat.hpp
@@ -34,7 +34,7 @@ class LP_TRANSFORMATIONS_API ConcatTransformation : public LayerTransformation {
     bool transform(TransformationContext& context, ngraph::pattern::Matcher &m) override;
     bool isPrecisionPreserved(std::shared_ptr<Node> layer) const noexcept override;
     bool canBeTransformed(const TransformationContext& context, std::shared_ptr<Node> layer) const override;
-    static bool isQuantizedStatic(const std::shared_ptr<const Node>& layer) noexcept;
+    static bool isQuantizedStatic(const std::shared_ptr<const Node>& layer);
 
 protected:
     static bool isHandled(
diff --git a/src/common/low_precision_transformations/include/low_precision/eltwise_base_transformation.hpp b/src/common/low_precision_transformations/include/low_precision/eltwise_base_transformation.hpp
index 189a24f704a7e0..80edca090e8256 100644
--- a/src/common/low_precision_transformations/include/low_precision/eltwise_base_transformation.hpp
+++ b/src/common/low_precision_transformations/include/low_precision/eltwise_base_transformation.hpp
@@ -22,7 +22,7 @@ class LP_TRANSFORMATIONS_API EltwiseBaseTransformation : public LayerTransformat
     bool canBeTransformed(const TransformationContext& context, std::shared_ptr<Node> layer) const override;
     bool isPrecisionPreserved(std::shared_ptr<Node> layer) const noexcept override;
 
-    static bool isBroadcasted(const PartialShape& shape) noexcept;
+    static bool isBroadcasted(const PartialShape& shape);
 protected:
     int getNotEmpty(const std::shared_ptr<Node>& eltwise) const;
     // Return indexes:
diff --git a/src/common/low_precision_transformations/include/low_precision/multiply_to_group_convolution.hpp b/src/common/low_precision_transformations/include/low_precision/multiply_to_group_convolution.hpp
index f31c26e424b1b2..9df09d0fccf1ed 100644
--- a/src/common/low_precision_transformations/include/low_precision/multiply_to_group_convolution.hpp
+++ b/src/common/low_precision_transformations/include/low_precision/multiply_to_group_convolution.hpp
@@ -31,8 +31,8 @@ class LP_TRANSFORMATIONS_API MultiplyToGroupConvolutionTransformation : public L
     bool transform(TransformationContext& context, ngraph::pattern::Matcher &m) override;
     bool canBeTransformed(const TransformationContext& context, std::shared_ptr<Node> layer) const override;
     bool isPrecisionPreserved(std::shared_ptr<Node> layer) const noexcept override;
-    bool isQuantized(const std::shared_ptr<const Node>& layer) const noexcept override;
-    static bool canBeTransformedToGroupConvolution(const std::shared_ptr<const Node>& layer) noexcept;
+    bool isQuantized(const std::shared_ptr<const Node>& layer) const override;
+    static bool canBeTransformedToGroupConvolution(const std::shared_ptr<const Node>& layer);
     static bool isDynamicOrScalar(const std::shared_ptr<const Node>& node);
 
     void setGroupSize(const size_t groupSize);
diff --git a/src/common/low_precision_transformations/src/concat.cpp b/src/common/low_precision_transformations/src/concat.cpp
index cdaf1ea3d34fed..498bf65bfec794 100644
--- a/src/common/low_precision_transformations/src/concat.cpp
+++ b/src/common/low_precision_transformations/src/concat.cpp
@@ -331,7 +331,7 @@ bool ConcatTransformation::isHandled(const TransformationContext& context, const
     return false;
 }
 
-bool ConcatTransformation::isQuantizedStatic(const std::shared_ptr<const Node>& layer) noexcept {
+bool ConcatTransformation::isQuantizedStatic(const std::shared_ptr<const Node>& layer) {
     const auto concat = as_type_ptr<const opset1::Concat>(layer);
     if (concat == nullptr) {
         return false;
@@ -342,8 +342,8 @@ bool ConcatTransformation::isQuantizedStatic(const std::shared_ptr<const Node>&
         return false;
     }
 
-    const size_t normalizedAxis = ngraph::normalize_axis(concat->get_friendly_name(), concat->get_axis(), outputRank);
-    return normalizedAxis == 1ul;
+    const auto normalizedAxis = ngraph::normalize_axis(concat->get_friendly_name(), concat->get_axis(), outputRank);
+    return normalizedAxis == 1;
 }
 
 } // namespace low_precision
diff --git a/src/common/low_precision_transformations/src/eltwise_base_transformation.cpp b/src/common/low_precision_transformations/src/eltwise_base_transformation.cpp
index 8b7140ce666b73..327075ba5fedab 100644
--- a/src/common/low_precision_transformations/src/eltwise_base_transformation.cpp
+++ b/src/common/low_precision_transformations/src/eltwise_base_transformation.cpp
@@ -15,7 +15,7 @@ using namespace ngraph;
 using namespace ngraph::pass;
 using namespace ngraph::pass::low_precision;
 
-bool EltwiseBaseTransformation::isBroadcasted(const PartialShape& shape) noexcept {
+bool EltwiseBaseTransformation::isBroadcasted(const PartialShape& shape) {
     const auto rank = shape.rank();
     if (rank.is_dynamic()) {
         return false;
diff --git a/src/common/low_precision_transformations/src/fake_quantize_dequantization.cpp b/src/common/low_precision_transformations/src/fake_quantize_dequantization.cpp
index 5bdec2f9e006b8..f1fff57fcf4ec0 100644
--- a/src/common/low_precision_transformations/src/fake_quantize_dequantization.cpp
+++ b/src/common/low_precision_transformations/src/fake_quantize_dequantization.cpp
@@ -75,7 +75,7 @@ bool FakeQuantizeDequantization::isLowPrecision() const {
     return DataPrecision::isSupported(data.get_element_type());
 }
 
-bool FakeQuantizeDequantization::checkShape(const std::shared_ptr<ngraph::Node>& elementwise) noexcept {
+bool FakeQuantizeDequantization::checkShape(const std::shared_ptr<ngraph::Node>& elementwise) {
     std::shared_ptr<ngraph::opset1::Convert> convert;
     std::shared_ptr<ngraph::opset1::Constant> constant;
     const int branchIndex = FakeQuantizeDequantization::fillDequantizationParams(elementwise, convert, constant);
@@ -90,7 +90,7 @@ bool FakeQuantizeDequantization::checkShape(const std::shared_ptr<ngraph::Node>&
     }
 
     if (!inPShape.rank().is_dynamic()) {
-        for (int i = 0; i < inPShape.rank().get_length(); ++i) {
+        for (int i = 0; i < inPShape.size(); ++i) {
             if (inPShape[i] != outPShape[i] && !inPShape[i].is_dynamic()) {
                 return false;
             }
@@ -178,7 +178,7 @@ std::shared_ptr<Node> FakeQuantizeDequantization::copyWithNewInput(const std::sh
 int FakeQuantizeDequantization::fillDequantizationParams(
     const std::shared_ptr<ngraph::Node>& elementwise,
     std::shared_ptr<ngraph::opset1::Convert>& convert,
-    std::shared_ptr<ngraph::opset1::Constant>& constant) noexcept {
+    std::shared_ptr<ngraph::opset1::Constant>& constant) {
     auto fill = [](
         const std::shared_ptr<ngraph::Node>& elementwise,
         const size_t branchIndex,
@@ -211,7 +211,7 @@ int FakeQuantizeDequantization::fillDequantizationParams(
 
 int FakeQuantizeDequantization::fillDequantizationParams(
     const std::shared_ptr<ngraph::Node>& elementwise,
-    std::shared_ptr<ngraph::opset1::Constant>& constant) noexcept {
+    std::shared_ptr<ngraph::opset1::Constant>& constant) {
     constant = elementwise->get_input_element_type(1ul).is_real() ?
         ov::as_type_ptr<opset1::Constant>(elementwise->get_input_node_shared_ptr(1ul)) :
         nullptr;
diff --git a/src/common/low_precision_transformations/src/multiply_to_group_convolution.cpp b/src/common/low_precision_transformations/src/multiply_to_group_convolution.cpp
index 8868c45f59a0cc..56acf72f21642e 100644
--- a/src/common/low_precision_transformations/src/multiply_to_group_convolution.cpp
+++ b/src/common/low_precision_transformations/src/multiply_to_group_convolution.cpp
@@ -201,11 +201,11 @@ bool MultiplyToGroupConvolutionTransformation::canBeTransformed(const Transforma
     return true;
 }
 
-bool MultiplyToGroupConvolutionTransformation::isQuantized(const std::shared_ptr<const Node>& layer) const noexcept {
+bool MultiplyToGroupConvolutionTransformation::isQuantized(const std::shared_ptr<const Node>& layer) const {
     return MultiplyToGroupConvolutionTransformation::canBeTransformedToGroupConvolution(layer);
 }
 
-bool MultiplyToGroupConvolutionTransformation::canBeTransformedToGroupConvolution(const std::shared_ptr<const Node>& layer) noexcept {
+bool MultiplyToGroupConvolutionTransformation::canBeTransformedToGroupConvolution(const std::shared_ptr<const Node>& layer) {
     const auto parent0 = layer->get_input_node_shared_ptr(0);
     const auto parent1 = layer->get_input_node_shared_ptr(1);
 
@@ -219,7 +219,7 @@ bool MultiplyToGroupConvolutionTransformation::canBeTransformedToGroupConvolutio
         return false;
     }
 
-    return (pShape.rank().get_length() == 4ul) || (pShape.rank().get_length() == 5ul);
+    return (pShape.size() == 4ul) || (pShape.size() == 5ul);
 }
 
 bool MultiplyToGroupConvolutionTransformation::isDynamicOrScalar(const std::shared_ptr<const Node>& node) {
diff --git a/src/common/low_precision_transformations/src/shuffle_channels.cpp b/src/common/low_precision_transformations/src/shuffle_channels.cpp
index f088be178e1529..0a46f013c9cf6c 100644
--- a/src/common/low_precision_transformations/src/shuffle_channels.cpp
+++ b/src/common/low_precision_transformations/src/shuffle_channels.cpp
@@ -92,6 +92,20 @@ bool ShuffleChannelsTransformation::canBeTransformed(const TransformationContext
         return false;
     }
 
+    // It's impossible to normalize a negative axis in case of dynamic rank
+    // but it's necessary when dequantization operations are per channel
+    if (shuffleChannels->get_input_partial_shape(0).rank().is_dynamic() && shuffleChannels->get_axis() < 0) {
+        const bool perChannelSub = dequantization.subtractConstant ?
+            ov::shape_size(dequantization.subtractConstant->get_shape()) > 0 :
+            false;
+        const bool perChannelMul = dequantization.multiplyConstant ?
+            ov::shape_size(dequantization.multiplyConstant->get_shape()) > 0 :
+            false;
+        if (perChannelMul || perChannelSub) {
+            return false;
+        }
+    }
+
     return true;
 }